tv Call-in with Thomas Malone Superminds CSPAN October 10, 2019 5:27am-5:46am EDT
we are back live at the national book festival. we're pleased to have join us now on our book tv set professor thomas malone of mit. here's his most recent book. it is called "super minds", the surprising power of people and computers, thinking together. before we get into the topic of the book, professor, what is it that you do at mit? >> i'm a professor in the sloan school of management and director of the mit center for collective intelligence. pleasure to be a here. >> what kind of management training do you give at mit? >> i teach two main courses, one is an mba course on strategic organizational design, about how to organize companies in different situations including innovative new things, like w i wikipedia for instance and the
other is a leadership workshop help students learn about different capabilities for leadership. >> what is the collective intelligence center? >> we study the kinds of things that i wrote about in my book. collective intelligence i define in a very general way as groups of individuals acting collectively in ways that seem intelligent. so by that definition, you could say every hierarchal company or nonprofit is a kind of collective intelligence, so are markets, so are communities, so are democracies. those are all examples of collective intelligence. and to foreshadow just a little bit what you may be about to ask me, i talked for a long time about these things as collectively intelligence systems, and i eventually realized that a good short way of saying that was superminds, so that's the name i chose for my most recent book.
>> how has our hyperconnectivity in the year 2019 change how you view this? >> well, from my point of view, hyperconnectivity is a very important part of the kinds of superminds that technology makes possible. i think many people are probably overestimating how important artificial intelligence will be. i think it will be very important, but i think people are underestimating how important hyperconnectivity will be. by hyperconnectivity, i just mean connecting people to other people and often to computers at huge new scales and in rich new ways that were never possible before. so the internet is the prime enabler for the hyperconnectivity that's happening all around us all the time, and i think we're just scratching the surface of what that's going to make possible. >> what do you see as being possible? >> well, wikipedia i think is an
interesting example. it's now happened already so we can understand it, but if you think about how an encyclopedia would have been written 30 years ago, you know, you would have had people sending letters or maybe some e-mails back and forth. you would have had editors. you would have had some world experts who wrote the things and reviewed the things. but because of the essentially free communication all over the world, now thousands of people have been -- people have been able to create an encyclopedia far larger than anything we have ever had before, in many ways, probably far better, and by the way, almost for free. that's a way that hyperconnectivity made it possible to organize the work of writing a very big document much more easily in a quite different way. >> you also say that we're overestimating the use or the
ability of artificial intelligence which is a huge buzz term right now. >> absolutely. i use the buzz term a lot. i teach on-line executive education courses, which i guess i should have mentioned earlier, about artificial intelligence. the one -- the first one we did is called artificial intelligence implications for business strategy. so lots of people are interested in this. i think there are huge potential for how artificial intelligence can be used in business and other parts of our society. but i think many people imagine that artificial intelligence will soon be kind of like people, doing the same things that people do, sitting in the driver's seat of a car, sitting in the boss's desk. i think that's a really misleading way of thinking about what artificial intelligence will enable to us do. -- us to do. >> today what are the most common uses of ai in our world?
>> so it depends of course on how you define ai. i would say google is a kind of artificial intelligence in a sense to search algorithms there. i would say that many companies are now using chat box and things like that that use some kinds of ai to communicate with people in something like natural language. there are more and more things like that. i think what will happen is that more and more of the small parts of what needs to be done in work will be done by artificial intelligence. but these ai programs today and for the foreseeable future are only capable of what i call specialized intelligence, doing particular tasks. they don't have the same kind of general intelligence that we humans do, the ability to talk about lots of different subjects, to use common sense, etc. i think that's quite a long ways
away probably. >> you say we're just scratching the surface on these areas; correct? >> yes -- both actually but i especially meant the surface of the possibilities for hyperconnectivity. >> let's go back to the overestimating of ai, though, in 10 years, 20 years, what is in your head? what do you see us doing or using this? >> let me tell you what people have estimated in the past as part of a way of answering that question. first, actually, i think the reason it's so easy to overestimate the potential of ai is because our science fiction is full of computers that are as intelligent as humans, but it is much easier to imagine such computers than it is to actually create them. if you ask people today how long it will take till we have human level artificial intelligence, an average answer would probably be about 20 years from now.
that's what many people would estimate. but what a lot of people don't know is that people have been asking that same question ever since the beginning of the field of artificial intelligence, in the 1950s, and people have estimated that human level ai has been about 20 years away for the last 60 years. so i think it is unlikely that we'll do that, that we will have that in the next couple of decades. >> we have a few more minutes with our guest thomas malone, we will put the numbers up in case you would like to participate in a conversation with the author of this book "superminds". go ahead and dial in. we'll begin taking those calls in just a minute. i think a lot -- i don't think a
lot -- i think about ibm's watson. do you have any idea how much has been invested in that program? have you had a role in it? and what is its function? >> so i haven't had any personal involvement with ibm watson. i do use it as an example, however. the original ibm watson program played the game show "jeopardy" and did it better than any human players of the game. i do know the person who led the development of that original watson program, and i confirmed with him what i suspected, which is that the original watson program, that was so good at playing jeopardy couldn't even play tic tac toe, much less chess. it was very specialized for that particular task. now ibm has used the term watson to describe other software they have done since then. but i think in some ways that may make it harder to understand
that many of those other programs are really just other programs that are called the same thing. and i think that we're still a long way from a general purpose ai program that can do all kinds of different things >> so you also teach an mba class. would you advise your students to invest and/or participate in the watson program? >> i don't have any particular recommendations about any company's products for or against. i think that the enterprise that ibm and many other companies that are engaged in, of bringing artificial intelligence techniques to bear on more and more problems in business and in medicine and so forth, i think that's a very worthwhile enterprise, one i would absolutely engage my students to engage in at ibm and many other places. >> back to hyperconnectivity, is it in a sense crowd sourcing? >> that's one example of what
you can do with hyperconnectivity. by crowd sourcing, people often mean letting anyone who wants to participate and that's one thing you can do very easily with hyperconnectivity, and there's been quite amazing things that have been done that way, getting good ideas from anyone, anywhere in the world about how to do these things. i think that's one, but not the only way of using hyperconnectivity. >> jean's calling in from maryland. hi, jean, you are on with thomas malo malone. >> good afternoon, dr. malone. i totally agree that having done 35 years of computer program, starting from 1958, i totally agree with your assessment of artificial intelligence which first popped up with very bright people trying to make things happen about 40 years ago. the results so far are interesting, but not really major, major -- [inaudible]. as far as people using computers, i must give an example. john hopkins did -- created the
first global satellite navigation system for the navy, with a team of four people doing the program, four on the navigation program. the point was in all the problems we had of every kind, from the science to the computers, the atmosphere at the lab was that everybody that had any idea when we hit -- we would have group meetings. anybody that could solve a problem, be our guest. that system ran operationally successful 19 years before gps. i thought you might be interested in an example. the way i translate what you are saying is a perfect example of the major factors you're talking about. thank you for your time. >> thank you, gene. >> great example, i think you described the team of people who work together on creating that computer program, it sounds like your team of people was a kind of supermind itself, and then
what they produced was a software program that could do a lot of other cool things. and i think you said artificial intelligence hasn't amounted to much yet. i think we are seeing more and more very useful things that artificial intelligence is doing. credit risk evaluation, for instance, with credit card charges and things like that. that's an example of something that happen so frequently we almost take it for granted, but that's a fairly early example of many of the benefits of artificial intelligence. >> next call for thomas malone comes from kathy in new york. hi, kathy. >> caller: hi, how are you? i watch your show all the time. >> thanks for watching. we appreciate it. what's your question? >> i wanted to ask dr. malone if hyperconnectivity could be related to emergent behavior in
groups of lower order life forms. in other words, like how colonies, like birds and ants and schools of fish, would he consider that a form of hyperconductivity? >> thank you, ma'am. so i think the example you bring up is a very interesting one. i would absolutely consider collections of animals as in many cases examples of collective intelligence or i'd call them superminds, birds flocking together, ant colonies, beehiv beehives, etc., those are all examples of what i think are interesting cases of collective intelligence. i wouldn't p usually -- in most cases i would not consider them examples of hyperconnectivity, because they still communicate with each other only at a fairly local level. so one ant can communicate with
another ant that's right next to it physically. one bird can sense the other bird flying near it. but they don't have as far as i know the kind of long-distance communication across now the entire planet that we humans have. and it's that kind of hyperconnectivity that i think will lead to some even more interesting kinds of collective intelligence and superminds. >> what's the down side to hyperconnectivity? >> well, it's a good question, and one we need to keep in mind. it is possible for people to know too much. i was just talking a few days ago with one of my colleagues at mit, who is a nobel prize winner in economics, and he was espousing the view that many things that we want to be able to do, for instance, in government, if we have too much connectivity, too much transparency, may make it harder to actually do what needs to be done.
if the people making decisions have to always worry about what anyone in the world would think about every comment they might make about what should be done, they are playing a game that's not just making the best decision. they're playing the game of how can i say things that people will be glad they heard me say? that doesn't necessarily lead to good decisions in the long run. >> one of the terms you use and talk about in your book "superminds" liquid democracy, which is what? >> liquid democracy is an interesting possibility that i think we don't yet know for sure that it will work on a large scale, but i think it is a very interesting possibility that we should be experimenting with much more seriously. the idea of liquid democracy is that you can have a kind of combination of direct democracy and representative democracy. in a direct democracy, each voter votes directly on every question. in a representative democracy, like we have in the u.s. and other countries today, you elect representatives who in turn vote on your behalf.
in a liquid democracy, you can do both. as an individual voter, you can always vote directly on any question, if you want to. but most of us don't begin to have the time or even the interest in doing that. so we can also in a liquid democracy delegate our right to vote, our proxy to anyone else we want to. i might give you for instance my proxy for voting on military issues. i would give my wife a proxy for voting on environmental issues. each of you in turn could delegate my proxy even further for instance for people who knew more about specific questions that needed to be voted on. if at any time i didn't like how my proxy was being voted, i could always take it back and either vote directly myself or give my proxy to someone else. see the potential advantage of that is it lets you create democracy that are i think often
much more response i have to the actual -- responsive to the actual desire of voters, better able to take advantage of specialized knowledge of particular questions, and the key point here is that this kind of democracy wouldn't even be feasible without computers. >> thomas malone is the author. he's at mit. his most recent book "superminds". previously the author of "the future of work" which came out